AD823 Lo pass filter schematic

[jmar]

There is a post in these archives that is titled "PCM2702 Usb-dac" and was started by member 00940.

I have a 2902 that works fine but I breadboarded his AD823 output filter.....have double and triple checked, and can't get it to function.

At best, I hear very faint audio.

Just as a time saver, would anyone know what might be the problem in the schematic?

I didn't see anyhting in the replies on the thread.

[danb1974]

Not sure I understand your problem.

2902 is fine, ie. audio signal before lpf is ok? If so, lpf opamp is correctly powered (v+, v-, vcomm)? Passive components checked with a multimeter?

If not, volume from windows mixer is at maximum?

[jmar]

All of the obvious are fine and my 2902 performs flawlessly.

I'm talking about the schematic on the post that I refer to.

I copied it EXACTLY and nothing.

[danb1974]

In the original TI schematics opamps + input is connected to pcm vcom pin and not to ground (ok they use another opamp but that should not make any difference)

Have a look at the lpf implementation from

http://focus.ti.com/lit/ug/sbau025/sbau025.pdf

This one works, have build 2 boards, no problems (except for a bad soldering I had to fix)

[jmar]

Dan....

Next time I'm in Timisoara, I'm buying all the Romanian beer you can take.

Damn....that was it, I just took the + pins to Vcom and it sounds great!
After 2 days of head scratching.

In fact, I'm really impressed with the sound of those AD823 opamps....much nicer than just the caps I had on the outputs.

Thanks so much for the heads up on that.....I owe you one.

Obviously the chip's analog audio ground doesn't come out with those
analog ground pins.

I'm an idiot for missing it but just wanted to save time by following that schematic.


THANKS !!

[danb1974]

Glad to hear it works. I'm impressed with the sound of my pcm2702 boards, tested on a decent system it was on par with a vincent cdplayer.

Not sure if it is ok to just skip the lpf as many do, I don't expect a series caps to filter the high frequency residues of the oversampling.

I'll trade the beer for a pic of your board. Mine is here (no box yet).

http://www.diyaudio.com/forums/showt...490#post826490

[jmar]

You know, I'm still puzzled why this happened and why I didn't get audio at first.

Maybe some others can look at this or I'll start a new thread.

For example....

(1) If you look at TI "typical circuit connection" in the 2902 datasheet, there is NO MENTION of a connection to Vcom.

(2) Why could I get audio signal by simply using a 10uf capacitor on the output AND GROUND? ! (NO connection to Vcom). Yet when I lifted the opamp pins from ground and went straight to Vcom, everything was fine?

(3) AND...I breadboarded an AD823 with the + pin going to ground, fed it audio and it worked fine.......hence, you can understand my CONFUSION.

Should pin14 (Vcom) be grounded? or left floating through the 10uf cap I wonder?

[danb1974]

The datasheet does not detail the post lpf implementation but it explains vcom pin. Look at the official dem-pcm2902 demo board and you can see there how vcom is used.

vcom carries VCCI/2 voltage and is "ground" for the dual-voltage-powered opamps. Don't know how to put it in english and very rusty about how an opamp works, you can look it on google, opamp v- is not ground but negative voltage from this ground and v+ is positive voltage from ground.

You could probably wire +in to a 1/2 resistive voltage divider connected between v+ and v- (done in opamp circuits in which no dual voltage supply available).

When you extract audio through a capacitor it filters the continous voltage so I think you should be able to use as the other wire (output ground) any stable voltage (be it ground or v+).